NOAA/WDS Paleoclimatology - Summer Air Temperature Reconstruction in Tierra del Fuego from 1765-2002 CE
收藏NOAA National Centers for Environmental Information2026-04-23 收录
下载链接:
https://www.ncei.noaa.gov/metadata/geoportal/rest/metadata/item/noaa-recon-37120/html
下载链接
链接失效反馈官方服务:
资源简介:
Proxy climate records, such as those derived from tree rings, are necessary to extend relatively short instrumental meteorological observations into the past. Tierra del Fuego is the most austral territory with forests in the world, situated close to the Antarctic Peninsula, which makes this region especially interesting for paleoclimatic research. However, high-quality, high-resolution summer temperature reconstruction are lacking in the region. In this study we used 63 tree-ring width chronologies of Nothofagus pumilio and Nothofagus betuloides and partial least squares regression (PLSR) to produce annually resolved December-to-February temperature reconstruction. The resulting reconstruction extends back to AD 1765 and explains 37–50% of instrumental temperature variability. We found that observed summer temperature variability in Tierra del Fuego is primarily driven by the fluctuations of atmospheric pressure systems both in the South Atlantic and South Pacific, while it is insignificantly correlated to major hemispheric modes: El Niño–Southern Oscillation and Southern Annular Mode. This fact makes our reconstruction important for climate modelling experiments, as it represents specific regional variability. Our reconstruction can be used for direct comparison with model outputs to better understand model limitations or to tune a model. The reconstruction can contribute to larger scale reconstructions based on paleoclimatic data assimilation. Moreover, we showed that PLSR has improved performance over principal component regression (PCR) in the case of multiple tree-ring predictors. According to these results, PLSR may be a preferable method over PCR for the use in automated tree-ring based reconstruction approaches, akin widely used point-by-point regression.



